Each year 230,000 people are hospitalized and survive moderate and severe traumatic brain injury (TBI). As a result, a large number of individuals with TBI endure life-long impairment and disability. Acute rating scales such as the Glasgow Coma Scale (GCS) have shown limited predictive validity regarding patient outcome and traditional neuroimaging techniques such as CT and MRI maintain limited correlations with brain injury severity and cognitive functioning. Continued advances in neuroimaging, however, have provided researchers with an important opportunity to study the pathophysiology of brain dysfunction following TBI. According to the NCMRR, "the neurobiology of TBI in humans should be studied with modern imaging techniques". The purpose of this study is to correlate proton magnetic resonance spectroscopy (MRS), an advanced neuroimaging technique, with behavioral measures of TBI severity and cognitive outcome. MRS measures the concentration of cerebral metabolites such as N-acetylaspartate (NAA), choline (Cho), and glutamate (Glu). While MRS has shown promise in predicting brain injury severity and patient outcome, the exact protocols for using MRS with TBI remain undetermined and the purpose of the proposed study is to examine three critical areas: (1) the post-injury time period when the MRS data should be acquired (e.g., within one week or within one month of injury); (2) how metabolites should be measured (i.e., absolute concentrations or changes in concentration over time); and (3) the brain locations best suited for MRS data acquisition (i.e., acquisition near lesion sites or acquisition at sites remote from probable brain lesion). The proposed study will make determinations in these three areas through the use of two acute MRS scans following TBI to measure concentrations of NAA, Cho and Glu and their correlation with injury severity and cognitive variables. In addition, correlation of acute MRS data with behavioral data (e.g., duration of loss of consciousness, duration of post-traumatic amnesia) will elucidate the relationship between changes in brain metabolism and changes in patient behavior during acute recovery from TBI. The present proposal will employ a promising, noninvasive neuroimaging technique, MRS, to determine the most appropriate protocols (i.e., timing, metabolic measurement, brain location for data acquisition) for application of MRS to acute TBI. With an established protocol for using MRS, this instrument should prove useful for determining the effectiveness of acute interventions (e.g. hypothermia, pharmacologic intervention) and for predicting the acute course of patient recovery.